基于JPEG-LS的多核压缩成像系统

打开文本图片集

关键词：图像压缩;CMOS探测器；现场可编程门阵列;流水线设计；多核压缩成像中图分类号：TP391.41；TP75 文献标识码：Adoi：10.37188/OPE.20263401.0139 CSTR：32169.14.OPE.20263401.0139

Multi-core compression imaging system based on JPEG-LS

LIU He，ZHANG Liu，LIGuiyang，SHEN Sangyuan，ZHANG Jiabao （ ， ， l3OOl2，） *Corresponding author，E-mail： zhangjiabao@jlu. edu.cn

Abstract： To address the constraints hardware cost and onboard data storage in high-resolution aero space CMOS cameras for smallsatellites，a full-pipeline FPGA-based implementation the JPEG-LS image compression algorithm is proposed. A multicore real-time compression imaging system is integrated on a single FPGA using a parallel grouping architecture. First，multichannel high-speed image streams from CMOS detectors are received by the FPGA. Second，the JPEG-LS algorithm is realized using an eleven-stage pipeline，and the encoding-parameter calculation and context-update modules are structurally optimized to reduce the critical path. Finally，multiple JPEG-LS compression kernels are employed to group and compress the multichannel data in paralel. Experimental results indicate that，，in the implemented system，the optimized JPEG-LS core achieves a maximum operating frequency 46MHz .With the compression parameter near set to 1，a near-lossess compression ratio exceeding 4 is obtained，and the peak signal-to-noise ratio （PSNR） the decompressed images is approximately 5O dB. These results satisfy the compresson-rate and image-quality requirements for remote-sensing imagery， providing a practical reference for designing high-resolution spaceborne CMOS cameras with integrated image compression.

Key Words： image compression; CMOS detector; Field Programmable Gate Array（FPGA）； pipeline; multi-core compression imaging

1引言

遥感卫星已广泛应用于环境监测、农业分析和城市建设等领域，随着遥感对地观测技术的不断进步，航天CMOS探测器的分辨率也不断提高，图像数据量以数倍增长，导致研发成本受限的航天遥感小卫星存储资源和星地下传资源紧张[1-2]。（剩余13315字）